justmars/statute-patterns

Construction of regex patterns for Philippine statutory law.

statute-patterns's Issues

Get the local path to the statute

• input: category and identifier
• output: local path

Serialize: Resolution of the Court En Banc - XX

Deal with legacy situation.

Bad codification event; see values={'locator': 'Section 1', 'caption': 'Scope.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991', 'action': 'Modified'}
Bad codification event; see values={'locator': 'A', 'caption': 'Pleadings allowed.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991'}
Bad codification event; see values={'locator': 'B', 'caption': 'Verifications.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991'}
Bad codification event; see values={'locator': 'Section 5', 'caption': 'Answer.', 'content': 'Within ten (10) days from service of summons,', 'statute': 'Resolution of the Court En Banc dated 10-15-1991'}
Bad codification event; see values={'locator': 'Section 6', 'caption': 'Effect of failure to answer.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991'}
Bad codification event; see values={'locator': 'Section 7', 'caption': 'Preliminary conference; appearance of parties.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991'}
Bad codification event; see values={'locator': 'Section 8', 'caption': 'Record of preliminary conference.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991'}
Bad codification event; see values={'locator': 'Section 9', 'caption': 'Submission of affidavits and position papers.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991'}
Bad codification event; see values={'locator': 'Section 10', 'caption': 'Rendition of judgment.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991'}
Bad codification event; see values={'locator': 'Section 18', 'caption': 'Referral to Lupon.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991'}
Bad codification event; see values={'locator': 'Section 19', 'caption': 'Prohibited pleadings and motions.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991'}
Bad codification event; see values={'locator': 'Section 20', 'caption': 'Affidavits.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991'}
Bad codification event; see values={'locator': 'Section 1', 'caption': 'Scope.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991', 'action': 'Modified', 'material_path': '1.1.2.7.9.3.', 'codification_id': 'mv-2022-rule-am-19-10-20-sc-civil-procedure-rules-of-court-v1'}
Bad codification event; see values={'locator': 'A', 'caption': 'Pleadings allowed.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991', 'action': 'Originated', 'material_path': '1.1.2.7.9.4.', 'codification_id': 'mv-2022-rule-am-19-10-20-sc-civil-procedure-rules-of-court-v1'}
Bad codification event; see values={'locator': 'B', 'caption': 'Verifications.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991', 'action': 'Originated', 'material_path': '1.1.2.7.9.4.', 'codification_id': 'mv-2022-rule-am-19-10-20-sc-civil-procedure-rules-of-court-v1'}
Bad codification event; see values={'locator': 'Section 5', 'caption': 'Answer.', 'content': 'Within ten (10) days from service of summons,', 'statute': 'Resolution of the Court En Banc dated 10-15-1991', 'action': 'Originated', 'material_path': '1.1.2.7.9.6.', 'codification_id': 'mv-2022-rule-am-19-10-20-sc-civil-procedure-rules-of-court-v1'}
Bad codification event; see values={'locator': 'Section 6', 'caption': 'Effect of failure to answer.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991', 'action': 'Originated', 'material_path': '1.1.2.7.9.7.', 'codification_id': 'mv-2022-rule-am-19-10-20-sc-civil-procedure-rules-of-court-v1'}
Bad codification event; see values={'locator': 'Section 7', 'caption': 'Preliminary conference; appearance of parties.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991', 'action': 'Originated', 'material_path': '1.1.2.7.9.8.', 'codification_id': 'mv-2022-rule-am-19-10-20-sc-civil-procedure-rules-of-court-v1'}
Bad codification event; see values={'locator': 'Section 8', 'caption': 'Record of preliminary conference.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991', 'action': 'Originated', 'material_path': '1.1.2.7.9.9.', 'codification_id': 'mv-2022-rule-am-19-10-20-sc-civil-procedure-rules-of-court-v1'}
Bad codification event; see values={'locator': 'Section 9', 'caption': 'Submission of affidavits and position papers.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991', 'action': 'Originated', 'material_path': '1.1.2.7.9.10.', 'codification_id': 'mv-2022-rule-am-19-10-20-sc-civil-procedure-rules-of-court-v1'}
Bad codification event; see values={'locator': 'Section 10', 'caption': 'Rendition of judgment.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991', 'action': 'Originated', 'material_path': '1.1.2.7.9.11.', 'codification_id': 'mv-2022-rule-am-19-10-20-sc-civil-procedure-rules-of-court-v1'}
Bad codification event; see values={'locator': 'Section 18', 'caption': 'Referral to Lupon.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991', 'action': 'Originated', 'material_path': '1.1.2.7.9.12.', 'codification_id': 'mv-2022-rule-am-19-10-20-sc-civil-procedure-rules-of-court-v1'}
Bad codification event; see values={'locator': 'Section 19', 'caption': 'Prohibited pleadings and motions.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991', 'action': 'Originated', 'material_path': '1.1.2.7.9.13.', 'codification_id': 'mv-2022-rule-am-19-10-20-sc-civil-procedure-rules-of-court-v1'}
Bad codification event; see values={'locator': 'Section 20', 'caption': 'Affidavits.', 'statute': 'Resolution of the Court En Banc dated 10-15-1991', 'action': 'Originated', 'material_path': '1.1.2.7.9.14.', 'codification_id': 'mv-2022-rule-am-19-10-20-sc-civil-procedure-rules-of-court-v1'}

Handle non-serialized rules

class Base(Enum):
    const = r"((PHIL\.\s+)?CONST(ITUTION|\.?)|(Phil\.\s+)?Const(itution|\.?))"  # can refer to 3
    admin = r"Administrative\s+Code"  # can refer to 3
    elect = r"Election\s+Code"  # can refer to 2
    corp = r"Corporation\s+Code"  # can refer to 2
    agrarian = r"Agrarian\s+Reform\s+Code"  # can refer to 2
    civil = r"Civil\s+Code"  # can refer to 2
    tax = r"(N\.?I\.?R\.?C\.?|National\s+Internal\s+Revenue\s+Code)"  # can refer to 3
    penal = r"Penal\s+Code"  # can refer to 2
    coop = r"Cooperative\s+Code"  # can refer to 2
    insure = r"Insurance\s+Code"  # can refer to 2
    rules = r"Rules\s+of\s+Court"  # can refer to 3

    @property
    def pattern(self) -> Pattern:
        return re.compile(self.value)

    def find(self, raw: str) -> Match | None:
        return self.pattern.search(raw)


NAMED_LAWS = [
    NamedLaw(
        base=Base.const.value,
        name="1987 Constitution",
        year=1987,
        serial=StatuteBase(
            statute_category=StatuteCategory.CONST, statute_serial_id="1987"
        ),
        aliases=[
            add_suffix_PH(rf"1987\s+{Base.const.value}"),
            rf"1987\s+{Base.const.value}",
        ],
    ),
    NamedLaw(
        base=Base.const.value,
        name="1973 Constitution",
        year=1973,
        serial=StatuteBase(
            statute_category=StatuteCategory.CONST, statute_serial_id="1973"
        ),
        aliases=[
            add_suffix_PH(rf"1973\s+{Base.const.value}"),
            rf"1973\s+{Base.const.value}",
        ],
    ),
...

Determine statutes found in a given text

from statute_utils import StatuteID
serial_title = 'Administrative Matter No. 05-8-26-SC'
l = StatuteID.get_statute_matches(serial_title.lower())

Serial Title reflects the path even in varianted cases.

In Administrative Matters, variant folder names become the id. This should not be the case since the serial title without the variant should be retained.

Add fields for sqlpyd use

class Rule(BaseModel):
    """Used to identify the statute based on a category and identifier. The `col` and `index` field attributes are not native to Pydantic; these are populated in anticipation of future use via the `sqlpyd` library."""

    cat: StatuteCategory = Field(
        col=str,
        index=True,
    )
    idx: constr(to_lower=True) = Field(  # type: ignore
        ...,
        regex="[a-z0-9-]+",
        col=str,
        index=True,
    )

This becomes usable in the following context:

class CodeRow(Page, Rule, TableConfig):  # see Rule use here
    __prefix__ = "lex"
    __tablename__ = "codifications"
    __indexes__ = [["statute_category", "statute_serial_id"]]
    statute_id: str | None = Field(
        None, col=str, fk=(StatuteRow.__tablename__, "id")
    )

TableConfig is sqlpyd-native and detects that the fields cat and idx are going to be part of the database table "codifications".

Bad logic in units creation

statute-patterns/statute_patterns/components/details.py

Line 31 in d3d997c

def set_units(cls, title: str, p: Path | None) -> list[dict]:

The appropriation markers in the title should only lead to a canned units result, i.e.

statute-patterns/statute_patterns/components/utils.py

Line 17 in d3d997c

UNITS_MONEY = [

If no existing units path exists. At present, it applies the canned unit results even with a pre-existing path.

Count number of unique statutes per text

class StatuteCounted(NamedTuple):
    statute: StatuteBase
    mentions: int

    @classmethod
    def extract_statutes(cls, content: str) -> Iterator[str]:
        """Extract iterator of statute serial texts so that these can be counted via `Counter()`"""
        if statutes := StatuteID.get_statute_matches(content):
            for s in statutes:
                if isinstance(s, StatuteBase):  # cat + idx exists
                    if s.is_single:
                        yield f"{s.statute_category} {s.statute_serial_id}"
                    else:  # more than one StatuteBase should exist
                        for dx in s.deconstructed_identifiers:
                            if isinstance(dx, StatuteBase):  # cat + idx exists
                                yield f"{dx.statute_category} {dx.statute_serial_id}"

    @classmethod
    def count_statutes(cls, content: str):
        statutes_detected = list(cls.extract_statutes(content))
        counted = Counter(statutes_detected)
        unique_statutes = iter(set(statutes_detected))
        for i in unique_statutes:
            mentions = counted[i]
            elements = i.split()
            cat = StatuteCategory(elements[0])
            idx = elements[1]
            try:
                base = StatuteBase(statute_category=cat, statute_serial_id=idx)
                yield cls(statute=base, mentions=mentions)
            except Exception as e:
                msg = f"Bad {elements=} in forming statute base; see {e=}"
                logger.error(msg)

Formula-based named patterns

Existing approach:

https://github.com/justmars/lawrgx/blob/8e893d0fa2c7a6b71059257971bbda887baf569a/lawrgx/names.py

Instead of manually constructing named patterns, may want to use something more programmatic. See experimental:

https://github.com/justmars/lawrgx/blob/8e893d0fa2c7a6b71059257971bbda887baf569a/lawrgx/recipes/base.py#L7

Need to refine digitrs

RA starts with 0:

CA/BP, etc. do not have the set_digits() function.

Get cat, idx from serial title

Node items have the format:

- locator: x
  action: y
  statute: Republic Act No. 386 # this is a standardized serial title

Extract the cat (ra) and idx (386) values from the serial title.

CircularSC and CircularOCA do not reflect proper serial title names.

Revised the serializer function.